AISEOCourse.academy Module 7

Module 07 · Lesson 4 of 4

Tying AI Visibility to Conversions

About 12 minutesPrerequisite: Lesson 7.3
After this lesson you canidentify AI-referral traffic in analytics, state exactly what attribution can and cannot prove, ask buyers where they heard of you, and present a stakeholder report that survives scrutiny.

Visibility that never becomes revenue is interesting data. Tying the two honestly is the difference between a measurement programme and a hobby, and honesty is the hard part, because AI search breaks the tidy referral chain that classic analytics was built on.

What you can see, and what you cannot

SignalWhere it showsHonest reading
AI-referral sessionsReferrer domains of assistants and AI search in analyticsReal but partial: app-based and logged-in use often arrives dark
Cited-then-searchedBranded search and direct traffic rising weeks after citation growthCorrelation with lag; suggestive, not proof
"Where did you hear of us?"Lead forms, sales calls, onboardingSmall sample, but the only direct evidence of the path
Landing-page conversion deltasPages you rebuilt in Module 4Confounded by everything else; use with baseline logic

The first row needs a caveat nailed to it: many AI surfaces send visits without a clean referrer, or from inside apps your analytics never labels. So treat AI-referral traffic as a floor, not a total, and say so in every report. The most underused row is the third: a mandatory free-text "where did you hear about us?" on lead forms costs nothing and, over a quarter, names the surfaces that actually feed the pipeline, including the assistant a dashboard cannot see.

DefinitionAttribution is the argument that a touch contributed to a conversion. In AI search the chain is usually unverifiable end to end; the honest version assembles several partial signals instead of asserting one line of causation.

The stakeholder report

One page, five blocks, every month:

  1. Visibility: the three shares and the benchmark table (lesson 7.3), with the denominator stated.
  2. Movement: what changed, tied to the work log (Module 6 cards shipped, Module 4 pages rebuilt) and to platform churn flags from the answer-changed column.
  3. Traffic and demand: AI-referral sessions (labelled a floor), branded search trend, direct trend.
  4. Pipeline: conversions, plus every "where did you hear of us?" answer that mentioned an assistant or AI search.
  5. Next: the top three opportunity cards and what they need.

The report's credibility comes from what it refuses to claim. "Citation share 12 to 19 percent over the quarter; AI-referral sessions are a floor at 340; four new customers named an assistant unprompted" is a paragraph a CFO can act on. "AI search drove 40 percent of growth" is a paragraph a CFO can destroy.

Worked example: reading a quarter honestly (illustrative)
  1. Citation share rose 7 points; two flipped queries match two Module 4 rewrites shipped in month one. Plausible contribution, noted as such.
  2. AI-referral sessions tripled to a floor of 340. Branded search up 18 percent with a three-week lag behind the citation growth. Consistent with cited-then-searched behaviour; labelled correlation.
  3. Two signed customers wrote "asked ChatGPT for X" in onboarding. Direct evidence, small n, quoted verbatim.
  4. Conclusion written: visibility growth is real; pipeline evidence is emerging, not established; continue the programme.
Workbench 7.4
  1. In analytics, find referrer domains from AI assistants and AI search; record this month's sessions and label the number a floor.
  2. Add "where did you hear about us?" (free text) to your lead form or onboarding call script this week.
  3. Draft the one-page monthly report with the five blocks, using your real numbers.
  4. Write your standard caveat sentence once, and paste it under the traffic block every month.
Self-check
Why is AI-referral traffic a floor rather than a total?
Because much AI-mediated use arrives without a clean referrer (in-app sessions, logged-in surfaces) and is invisible or mislabelled in analytics. You see part of the flow, so you report a minimum, never the whole.
Branded search rose after citation growth. Did the citations cause it?
Plausibly, with a lag: cited-then-searched is a documented behaviour pattern. But it is correlation. Report it as consistent-with, alongside any direct evidence from the "where did you hear" answers.
What makes the monthly report credible?
Stated denominators, labelled floors and correlations, verbatim customer evidence, and claims sized to what the evidence supports. The refusal to overclaim is the persuasive part.
Key principleAssemble partial signals and refuse the single causal line: honest attribution is what makes visibility reportable as business.

Sources used in this lesson
Google Search Central: AI features and your website
Google Search Central: SEO starter guide

Previous 7.3 Source Share and Benchmarks